Hi Kris,

I don't have any experience in thermocompression bonding or
electroplated gold, but I have done some work with thermally deposited
gold for MEMS applications. In my devices, the stress in the gold
layer changes significantly after a 200C bake.

I'd say that even if you manage to deposit your gold with zero stress,
as soon as you heat it beyond some critical temperature (~100C if the
gold is on silicon), it will become more tensile when you cool it back
to room temperature. The reasoning is explained well in "Creep of thin
film Au on bimaterial Au/Si microcantilevers" by K. Gall et. al..

Jason Milne
Microelectronics Research Group
The University of Western Australia


Quoting kris :

> Dear all,
>
> I am bonding (Au-Au thermocompression bonding)two
> different dies, one of them has electroplated Au on
> it. The sacrificial layer removal of one die results
> in the device release over the other die.
>
> I have seen lot of stress in the released structure. I
> was wondering if the the bonding temperature (400C) is
> the main reason for the residual stress in the
> electroplated Au??
>
> Can someone suggest me or provide a reference for the
> stress free or less residual stress Au electroplating.
>
> I am using sulphite bath from Technic, Inc. My plating
> setup is a typical beaker,electromagnetic stirrer,
> platinum mesh and 4" wafer. I am using pulsed power
> supply for the electroplating.